APBRmetrics

I'm sure stuff like this has been done many times before, but in any case...

Playoffs. 4th Quarter or any Overtime. 1 minute left in the period. Consider Field Goal Attempts to tie or take the lead.*

Data from Playoffs 2001-2012 (through May 29th 2012).

Restrict attention to players who’ve made the All-NBA First Team at least twice since the 2002-2003 season.
Restrict attention to players who attempted at least 10 such field goals.

We’re left with 6 players. In alphabetical order:
Kobe Bryant. Tim Duncan. Lebron James. Steve Nash. Dirk Nowitzki. Dwyane Wade.

From these 6 players, here are the FG%’s, in descending order:
50.0%
42.1%
36.8%
30.8%
27.0%
23.1%

Which player shot which %?

Just for fun, before I tell you the results, I want you to match players to the %. Click on this:
http://www.surveymonkey.com/s/MJ5QL6Y

Answers are revealed after you’ve tried it.

*That is, any FGA when it’s a tie game, down 1, or down 2; OR any 3FGA when down 3.

I didn't see any answers.

Answer was given at this page
https://docs.google.com/document/d/1l3c ... E0luw/edit

Please try out that surveymonkey thing before looking at the answer, I want to get a sense of what people's guesses are. Thank you!

OK, I got one right (I think), and it was Kobe, the only one with more than 20 FGA in the sample. He has had 37.
My other guesses were pretty random, relative to the facts. But these are based on as few as 13 shots.

~20 shots over however many post-seasons makes for small sample theatre... using minutes left and point margin thresholds to get at clutch just illustrates how trivial and unserious arguments about clutchness are, as the names change every time you change the thresholds or sample.

It would also be good to know about 3-pt shots vs 2-pt shots. When you need a 3 to tie, it may be that 23% is pretty good, relative to most players when the D is geared for stopping the 3.

Then figure out some way this doesn't reduce sample size even further.

Getting to the FT line, and making one's FT, you could have bigger samples, and still with some indication of "clutch".

Cool idea. Extremely small sample size of course, but it just goes to show that LeBron does not necessarily suck in the clutchiest of times.

Thinking about it some more, it seems you have to figure FT into this mix. The 23% FG shooter always takes it to the basket. The 50% shooter always shoots a fallaway. One guy is very confident (and I believe successful) at making clutch FT, and the other hates the very thought.

That's oversimplified, but it seems pretty relevant.

Making a 3, when a 3 is needed to tie, should perhaps be counted twice. That 3rd point is as vital as the first 2, and at least twice as challenging.
A made 3 to win, when a 2 would tie: count it 1.5 times. Didn't Duncan hit such a shot vs LA ?
A made 3 when a 2 would win it: should count no more than a made 2.

Thanks for the helpful comments, I shall try to incorporate your ideas into further analyses when I have time.

I do not think this little list I have produced is a definitive ranking of the clutchest superstars.

I started this little exercise, only as an attempt to convince a friend who bought into the myth that Kobe is one of the clutchest players and Lebron one of the chokiest. First I referred him to e.g. http://www.82games.com/1112/CSORT6.HTM
There we see that season after season Kobe is dominated by Lebron in most categories.

Then he said let's look at the last 1 minute, playoffs. So this is what I came up with.

Now he's asking me to look at FT's, and I think the same result will show.

So I think pretty much however you slice it and dice it, change the thresholds, change it to last 2 minutes or last 24 seconds, etc., Lebron is always going to "out-clutch" Kobe.

And that's all I wanted to prove to him.

The more I look at "clutch" stats, the more I wonder about their usefulness. The sample size is small, and the definition is arbitrary. "Clutch" analysis typically focuses on the end of close games, but in close games there are clutch plays made throughout the game.

Also, the analysis typically focuses mostly on shooting when there are LOTS of ways to make clutch plays that don't involve taking a shot.

There are many clutch opportunities throughout a typical game. Maybe the best clutch players are simply the best players -- the guys who make good plays all game long.

The biggest sample of "clutch" performance is probably just playoff performance.
Think Robert Horry vs Clifford Robinson.

Someone I was talking to said he would like to see these numbers expanded to the final 3 minutes. Is there any way you could provide breakdowns for multiple minute ranges? I know it might be a lot of work so I feel bad asking people here to compile things haha.

Nevermind, I found out how to do it myself.